무인항공기 이동 목표물 추적 비행 및 착륙 성능 개선을 위한 입력 신호 변형 기반의 제어 보조 시스템

Abstract

In this study, a novel input signal transformer is introduced to remove an overshoot and steady-state error that hinder stable tracking and landing of an unmanned aerial vehicle (UAV) based on image recognition, and the performance isverified. The Input Signal Transformer (IST) is designed in the shape of a sigmoid function to attenuate or amplify the input signal to resolve the aforementioned problems. For the verification of the UAV control system based on the IST, the UAV, target, and sensors were implemented in a virtual environment using the robot operating system (ROS). And data exchange structure and control system were built by the ROS-based message communication. A simulation was performed to confirm the elimination of the overshoot and steady-state error when the UAV to which the above control system was applied tracks and lands a fixed target and a moving target. As a result of the simulation, when the IST was not used, the UAV performed an unstable because of the overshoot and could not land on the target by the steady-state error. Conversely, in the case of using IST, it was confirmed that the flight was stable and landed successfully.